The brain's ability to efficiently carry out a task relies not only on the engagement of cognitive processes relevant to the task at hand, but also on the suppression of task-irrelevant cognitive processes. The inability to suppress task-irrelevant processes not only drives lapses of attention in everyday life, but also is characteristic of many neuropsychiatric disorders. For example, many patients with severe depression are unable to turn off negative ruminations. Neuroimaging work has identified distinct brain networks engaged in externally- versus internally-directed cognition, which exhibit opposite patterns of activity during both task and rest conditions. A default-mode network (DMN) activates when subjects engage in self-referential processes such as remembering a past event or planning for the future. The DMN also deactivates when subjects engage with the external environment, and decreased deactivation is associated with worse behavioral performance on such tasks. Conversely, a dorsal attention network (DAN) activates when subjects attend to stimuli in the environment. Control networks, which are thought to configure more specialized brain networks in order to carry out a task, have been shown to selectively link with task-relevant brain regions; for example, with the DMN when retrieving a memory, or with the DAN during visual search. However, the mechanisms of task- related deactivations are still unknown; for example whether they result from direct mutual inhibition between task-relevant and -irrelevant brain regions, or are mediated by a third set of regions such as the control networks. Moreover, while correlative evidence has linked the deactivation of task-irrelevant brain regions to better behavioral performance, there is no causal evidence in support of this idea. The proposed work aims to address these gaps by using electrocorticography (ECoG) to record electrical activity directly from human cortex with high temporal and anatomic precision, from regions previously shown by our lab to activate during memory (internal cognition) versus math processing (external cognition). We will track the dynamic interactions between, and relative timing of activity in, math-selective and memory-selective brain regions, as well as control-regions, during math versus memory processes (Aim 1). We will also use electrical brain stimulation (EBS) to transiently perturb activity in the above regions during math versus memory processes, to assess the causal function of task-related deactivations on behavior, as well as on the neural activity in task-relevant brain regions (Aim 2). This work has the promise to provide the training ground best suited for my desired career in systems neurosciences and electrophysiology, and the resulting knowledge may elucidate the dysfunctional neuronal communication thought to underlie many neuropsychiatric disorders.